This study deals with the spatial and temporal changes in the composition of oxygen and hydrogen stable isotopes of precipitation on two small Japanese islands. In Miyake Island, seasonal changes in the isotopic composition plotted on a δ-diagram exhibit a distorted circle in the counterclockwise direction. Due to the detection of a similar trend in Tokyo, it is suggested that both areas are under the same air mass. Notably, a decrease in the δ-value in winter was identified in both areas. An analysis using the Rayleigh model suggests that snowfall on the Sea of Japan side influences the δ-values of Tokyo and Miyake Island. In contrast, a small seasonal change was observed in Hachijo Island; inflow of a different vapor mass on the island is suggested. The results of calculation of isotopic composition indicate that the vapor mass may have its origin in the sea area near Hachijo Island.

View full abstract

Yokota-Gire is the historic flood of the Shinano River which occurred in 1896 and the estimation of the magnitude of this flood was reported by the authors. In the Shinano River, the flood discharge from 1956 to the present has been measured and known, while for the period from 1897 to 1955 the flood discharge data is not known. What is known in this period is the flood stage data and the daily rainfall data of the basin. This study attempts to estimate the flood discharge for this period and obtain the annual maximum flood discharge data from 1896 to the present. The flood events from 1897 to 1964 are first classified as either rain-front-caused or typhoon-caused and their daily rainfall data are, by estimation, transformed to the hourly data, which then are processed through the runoff analysis of the Shinano River to yield the annual maximum flood discharge data. The end result of this process is the consecutive 105 annual maximum flood discharge data from 1896 to 2001. This set of data are then used as the input data for the statistical analysis to give the statistical evaluation of the Yokota-Gire flood discharge, which turns out to be of the magnitude of 1 over 120 to 350 years in the exceedance probability.

View full abstract

Oxygen and hydrogen isotopic ratios and Cl^- and SO₄^2- concentrations of rain and river water were measured for the purpose of understanding the discharge process in a river source region, and the following observation results were obtained. The isotopic ratios and the chemical concentrations are higher in forest than outof forest, implying evaporation on leaves and wash out the chemicals sticking to leaves. River water keeps fairly constant isotopic ratios and chemical concentrations. This means that a reservoir of groundwater discharging the river is large and homogeneous. Exceptionally, during snow melting as well as continuous heavy rainfall, the river water is affected, especially in the chemical concentration, by discharge of the snow melt water, or those rain water. In estimation of the direct runoff rate of rain water based on variations in the isotopic ratios and the chemical concentration of river water, the estimate rates are different between bases of the isotopic ratio and the chemical concentration, and beside between Cl^- and SO₄^2- concentration during a decreasing of the rate after the highest rate. These imply that the isotopic ratios and the chemical concentration of base flow groundwater or direct runoff rain water are not always the same as the values of river water before the beginning or the average value of rain water outof forest.

View full abstract

We found a high concentration of dissolved aluminum accompanied by almost neutral pH in stream water within a suburban secondary forest underlain by granite bedrock. We aimed to clarify the characteristics of the high Al levels in runoff by observing the chemical properties of spring water, stream water and surface soils. The mean pH values of spring water and stream water were 6.09 and 6.44, and the mean concentrations of dissolved Al of each after filtration with 0.45 μm membrane were 66 μmol/L and 41 μmol/L, respectively. The concentrations of dissolved Al in these waters were extremely high for such pH levels. There was no relationship between dissolved Al concentration and dissolved organic carbon (DOC), which, if it did exist, would be considered an indicator of organic species of Al in the stream water. The concentration of dissolved Al was highly correlated with Fe concentration but not with Si. The total concentration of water-soluble Al in soil extracts from the surface soils around the stream was extremely high, and the concentrations of dissolved Al, Fe and Si were highly correlated. These results suggest that the high concentration of Al in the spring water and stream water are derived from weathering of aluminosilicates and that Al might be dissolving in water as inorganic colloidal species.

View full abstract

In this study, the effect of integration of river model and estuary model was evaluated to demonstrate the benefit of an integrated estuary model. An integrated estuary model consits of hydrological model, estuary model and meteorological model. In general, salinity distribution may be found to be significant for the evaluation of long-term water quality in an enclosed bay because salinity mainly controls density in a bay. Therefore, the influence of tide on salinity in a bay was investigated to demonstrate the necessity of an integrated estuary model. Two different river discharges and three different tidal amplitudes were used to examine the influence of tidal effect on salinity flux at the mouth of river. It was revealed that, for the case where river discharge at the upstream end of river model is constant and river discharge due to tide is included, with increasing reduction of average salinity from initial conditions, the river discharge at the upstream end was smaller than river discharge due to tide. The integration of river model, which can include the feature of estuary, and estuary model is, thus, found to be important to evaluate salinity flux at the mouth of river accurately.

View full abstract

Hydrodynamic dispersion coefficients, D, for an aggregated Andisol were determined for saturated and unsaturated flow conditions. The dispersivities, λ, as a function of pore-water velocities, v, and volumetric water contents, θ, were compared with those for a non-aggregated dune sand. The water retention curve for the Andisol had a stepwise shape. The water content decreased drastically at air-entry h=-15 cm and lower pressure head h=-3,160 cm. The radius of the aggregate was assumed to be approximately 0.1 mm based on the air-entry value. The estimated θ for the aggregate were 0.5 cm³ cm^-3 while the saturated water content was 0.74 cm³ cm^-3. The dispersivity, λ, increased linearly with v and reached more than 2 cm for saturated water flux, q, ranging from 146 to 3,085 cm d^-1. This velocity dependency was due to solute exchange between inter-and intra-aggregate pores. The convection-dispersion equation (CDE) agreed well with measured breakthrough curves (BTCs) regardless of q values because of the relatively rapid solute exchange. On the other hand, λ for unsaturated conditions decreased as θ decreased from saturation, and had a constant value of 0.2 cm for θ=0.5-0.6 cm³ cm^-3. Effects of water flow in inter-aggregate pores decreased and flow in intra-aggregate pores became dominant for lower θ. The CDE also agreed well with unsaturated BTCs because of relatively homogeneous flow in the soil aggregates.

View full abstract

Snow water equivalent (SWE) increases linearly against the elevation in mountains below the forest limit, but this effect is not quantitatively evaluated when it is extrapolated above the forest limit. In this study, snow surveys were carried out not only below the forest limit but also above there. The study area was Mt. Makihata in Niigata prefecture during the springs in 2002 and 2003. SWE increased linearly against the elevation below the forest limit, but it decreased rapidly above there. The distributions of SWE in the basin were calculated and compared, based on the two kinds of altitudinal distribution of SWE above the forest limit; one was the traditional method that SWE was extrapolated from the distribution below the forest limit, and the other was based on the snow surveys above there. The basin of Uono river above Muikamachi was extracted from the digital elevation model, and the areal distribution of snow-cover in the basin was calculated from remote sensing analyses. The differences of SWE in the basin could be neglected in lumped models, because the differences were 8-12 % on average in the basin, being almost same as the observational errors of the precipitation or the river discharge. In distributed models, the differences of SWE were up to 300 % in the ridges of the mountains. In addition, the imbalance of wind-induced snow movement among basins was indicated, because the area above the forest limit in the basin has more slopes of windward than those of leeward.

View full abstract

In order to investigate the quality and formation of stream water, a fixed-point observation of water quality was conducted over two years in water conservation forest of the Kanto Mountains, which have similar land coverage and different geologic conditions. Factors that may affect the nature and quality of stream water of the studied area, such as geology, atmospheric deposition, vegetation, and climate, were discussed. The results of 470 water samples taken from 44 different streams showed that the tap water taken from the streams reflected the high levels of Ca₂⁺ and HCO₃^-. We also examined the relationship between water and the rock and mineral content in various soils. By studying the composition of predominant water samples, we were able to show a geochemical perspective. To try and discover the reason for the NO₃^- and Cl^- concentration in the water samples that were unexplainable from geology, we looked at atmospheric precipitation, vegetation and climate and found that the majority of the Cl^- could be attributed to atmospheric precipitation. Although only minor, our results proved that the indirect fertilizing effect that limestone had on the soil accounted for the NO₃^- concentration and although we were unable to find definite proof our results have lead us to believe that temperature could also be largely responsible for NO₃^- levels in the water samples. We believe that there are other factors that may have a significant effect on water quality such as “different vegetation”, i.e., geographical conditions, and surrounding vegetation and that there is a necessity for studies concentrating on these issues.

View full abstract

In Japan, forested area is generally distributed at higher elevation than other land-use such as agricultural area and urban area. Several observational studies have reported that precipitation amounts increase with elevation. These two facts imply that forested area has greater precipitation amounts than other land-use. However, no studies have quantified the difference in precipitation amounts between forest and other land-use such as agricultural area and urban area. The purpose of this study is to quantify the difference by using GIS datasets.
Annual precipitation for forested area, agricultural area, and urban area was 1900.7 mm, 1565.7 mm, and 1575.3 mm, respectively. The average of annual precipitation amounts for forested area was greater than those for agricultural area and urban area by ca. 330mmy-1. This difference was caused by following two facts, as expected before. First, forested area is generally distributed at higher elevation than other land- use. Second, precipitation amounts increase with elevation. Therefore, we should consider this difference, when comparing runoff and/or evapotranspiration between forested area and other land-use.

View full abstract

Although regionalization, regional flood frequency analysis and scaling in hydrology have been studied by many researchers in the United States, Europe and other countries in order to predict flood for long time after 1950's, these have not become general theories and not been developed, in Japan. Moreover, for example, a figure which shows the relationship between the mean of annual maximum discharge and catchment area, which are most valuable and popular figure in this research field, is sometimes regarded as a kind of so-called Creager curves which is theory on envelope curves of the maximum flood discharges and is not very useful for frequency analysis. However, these theories which we interpret are quite useful for estimating flood discharge (quantiles) by synthesizing data at plural observation stations. Moreover, these are quite valuable in terms of transferring data from a gauged basin to an ungauged basin, and these theories would be potential solution for flood frequency problem in poorly gauged/ungauged basins. One framework of scaling, namely, simple scaling is quite easy to understand intuitively. On the other hand, multiscaling is more mathematical and would be relatively hard to understand. In this paper, the authors intend to interpret the basic theory on both frameworks, and mention a perspective of flood frequency estimation in ungauged basins by referring regression analysis method which is developed in the United States.

View full abstract

“Where does water go when it rains on hillslope?” Researchers have worked on this question since studies by Horton in 1930s. Studies by forest hydrologists like Tsukamoto and Hewlett in the 1950s and 60s in forested upland catchments presented an alternative to infiltration excess theory in the form of the variable source area theory (VSA). This theory has been the basis for many topographically based conceptual rainfall-runoff models used today. Recent observations and monitoring of natural hillslopes in many areas of the Pacific Rim have begun to observe that VSA does not adequately explain the types of behaviors now uncovered through combined use of hydrometric, isotopic and geochemical measurements: threshold subsurface stormflow initiation, hysteresis in the hillslope-catchment discharge relation, dominance of vertical and lateral preferential flow and rapid mobilization of old water. Based on McDonnell (2003, Hydrological Processes, 17, pp. 1869-1875), this commentary summarizes advances in hillslope hydrology since 1960's, and examines the applicability of several VSA assumptions used in our current rainfall-runoff models. Further, we present recent challenges to introduce qualitative field information to rainfall-runoff models. Then, we try to summarize some possible ways forward to for how we as a community might develop a “post” variable source area concept.

View full abstract